210 E. A. ANDREWS 



a second from above crossed and 2 more crossed, but yet the 

 trail was not reestablished. The crowd that had accumulated 

 above surged back 3 inches to an old side path and soon ran in 

 it 5 or 6 abreast. Meantime some of them going along the 

 upper edge of the paper got 5 J inches to the end of it and so 

 around and back to the trail below, a soldier in the lead. At 

 the end of 1 5 minutes there was a strong line descending around 

 the paper and so by the old path to the ground, while but few 

 crossed the paper and these checked as they came to the edge 

 which was raised so the head of a' termite might go in under it. 

 But the next evening a slender stream of termites was going down 

 the tree across this paper, and the edge was chewed somewhat. 

 When running on a board the termites also bite off some of 

 the roughness left on the sawed surface and so smooth their 

 paths. 



A pencil mark across the trail did not check the march but 

 scraping off the surface of the board did. 



A like strip of paper 4 mm. wide was put across a strong 

 procession down the tree. After i minute of check a worker 

 ventured across and a second and third followed, but several 

 hundred crowded together at the upper edge and remained 

 while many ran up and down across the narrow strip which 

 is not the length of the smallest termite in the procession. But 

 as soon as the paper is removed the crowd flows down like a 

 liquid mass. When the same paper was moistened with saliva 

 and crossed, above the former place, the termites hesitated 

 riiuch more and few crossed for many minutes. 



The same paper pressed down across a trail on the sand 

 caused scarce an instant's pause till the termites walked across . 

 But at this same time, 9.45 p. m. scratching off the sand, even 

 removing the white sand down to the dark earth, caused but 

 little check and the termites soon crossed the new material, 

 though on previous evenings at 8.30 they had been much more 

 disturbed by such breaking of the trail. Apparently on the 

 sand they may be more influenced by general sense of direction 

 than on the tree and less by response to the trail and that at 

 different periods of the daily rhythms of activity the relative 

 intensity of these two factors may be different on the sand. 



The daily rhythms have also a bearing upon the above 

 recorded accumulations of termites above and below an obstacle. 



